Qualcomm inaugurated its Hsinchu building on March 17, and TSMC’s Senior Vice President of Business and Research Development for Europe and Asia, Hou Yongqing, attended the ceremony to pay tribute and participated in the industry summit hosted by Qualcomm, demonstrating the close relationship between the two companies.

As for the media’s concern about whether Qualcomm is evaluating film production at TSMC’s Arizona plant, Chen Ruowen said that Qualcomm began evaluating long ago and that Qualcomm would be the first customer for TSMC’s U.S. plant’s 4nm process.

As previously reported last December TSMC tripled its planned investment in the Arizona chip plant, which began construction late last year, to $40 billion (currently about 275.6 billion yuan). The plant, one of the largest overseas investments in U.S. history, is set to start production in 2026 with advanced 3nm technology.

However, Digitimes quoted sources as saying that, judging from the current engineering and equipment installation progress, TSMC’s new U.S. fab is unlikely to be fully operational in 2024, and is likely to be delayed until 2025.

A semiconductor equipment supply chain source revealed that the overall construction and equipment installation progress of TSMC’s new fab in Arizona had been delayed, while the foundry must also address severe manpower shortages, soaring costs and emerging education and adaptation issues involving foreign employees.

The post TSMC U.S. Arizona plant 4nm mass production next year, Qualcomm is the first customers appeared first on TechGoing.

It is reported that the new Snapdragon 7 series chips will use TSMC’s 4nm process because Samsung’s previous 4nm LPE process for Snapdragon 7 Gen1 chips is not as efficient as TSMC. From the perspective of energy saving, Qualcomm did not sign an OEM contract with Samsung for the new Snapdragon 7 series chips.

According to previous news, the new chip that Snapdragon will release this time is likely to be SM7475, which uses an octa-core CPU of 12.95Ghz+32.5Ghz+4*1.79Ghz, Adreno 730 GPU, and the AnTuTu comprehensive running score reaches 1029731 points, more than Dimensity 8200. The Geekbench 5 running score of the SM7475 chip has a single-core score of 1232 points and a multi-core score of 4095 points (This running score is higher than Dimensity 8200 and comparable to Dimensity 9000).

The post Samsung failed to sign contract to manufacture new Snapdragon 7-series chips appeared first on TechGoing.

The Snapdragon 8 Gen 2 chip has a peak frequency of 3.2GHz and the Snapdragon 8 Gen 2 overclocked version in the Samsung Galaxy S23 series has a peak clock speed of 3.36GHz. The Snapdragon 8 Gen 3 chip expected to be launched this year will bring better performance while reducing power consumption.

Snapdragon 8 flagship chips are usually launched in Q4 every year, previously in early December, but last year it was brought forward to November 16. According to Weibo blogger @digitalchat, the Snapdragon 8 Gen 3 chip will be released a little earlier this year (expected to be in late October or early November 2023). But the phone schedule for each manufacturer is still Q4.

The details of the Snapdragon 8 Gen 3 chip are not yet clear. According to the revealed scores, the Snapdragon 8 Gen 3 chip will have 25% better CPU performance than the Snapdragon 8 series SoCs. The Snapdragon 8 Gen 3 for Galaxy chip has a single-core score of 1,930 and a multi-core score of 6,236 in early Geekbench tests.

The word is that Snapdragon 8 Gen 3 for Galaxy will feature a 1+5+2 core configuration, one more power-efficient core than the current Snapdragon 8 Gen 2 for Galaxy’s 1+4+3 setup. Thanks to TSMC’s N4P process node, it could be up to 20 percent more energy efficient than Snapdragon 8 Gen 2.

The post Qualcomm Snapdragon 8 Gen 3 chip will be released earlier this year appeared first on TechGoing.

The MediaTek Dimensity 7200 uses the second-generation TSMC 4nm process, the same as the Dimensity 9200 series. Equipped with two Cortex-A715 cores with a peak frequency of 2.8GHz and six Cortex-A510 cores (frequency unknown). It is also supplemented by the Mali G610 MC4 GPU, which is similar to the G710 in the Dimensity 9000, but with fewer shader cores, can support FHD displays with a refresh rate of up to 144Hz, and also supports HDR10+, CUVA HDR and Dolby HDR . Powered by MediaTek’s HyperEngine 5.0, which allows for AI-based variable-rate shading, it also supports UFS 3.1 storage.

MediaTek said that this chipset is equipped with a 14-bit HDR ISP, supports 4K HDR video capture, and also supports a main camera of up to 200 million pixels. It’s even possible to receive two video streams, each in full HD. Thanks to the built-in APU, the processor also supports some AI-powered camera enhancements like real-time portrait enhancements.

The MediaTek Dimensity 7200 supports sub-6GHz 5G networks with a downlink rate of up to 4.7 Gbps. It supports 2CC carrier aggregation and dual 5G SIM, while also supporting tri-band Wi-Fi 6E and Bluetooth 5.3.

Dimensity 7200 chipset will be unveiled in 5G smartphones released globally in the first quarter of 2023.

The post MediaTek Dimensity 7200 processor released, using second-generation TSMC 4nm process appeared first on TechGoing.

Although theoretically overclocking is a good thing, this exclusive custom version has caused some users to worry, whether it is TSMC OEM or Samsung OEM has become the focus of the conversation.

Today, Qualcomm finally responded positively to user questions, saying that this custom version of Snapdragon 8 Gen 2 is still using TSMC 4nm process, rather than Samsung’s own OEM production.

This answer, no doubt, allows users who have experienced serious heat problems caused by the Snapdragon 888 and Snapdragon 8 Gen 1 Samsung OEM, to breathe a long sigh of relief.

The S23’s custom Snapdragon 8 Gen 2 CPU is said to be clocked at 3.36GHz, higher than the 3.2GHz standard version, and the GPU frequency is 719MHz, also higher than the 680MHz of the standard version.

This gives the Galaxy S23 series a Geekbench 5 single-core score of 1583, compared to the standard version of the second-generation Snapdragon 8 with a single-core score of around 1480 and the MediaTek Dimensity 9200 with a single-core score of between 1400-1500.

The post Qualcomm confirms Samsung Galaxy S23 overclocked version of Snapdragon 8 Gen2 is still TSMC 4nm process appeared first on TechGoing.

Then, another familiar digital blogger @digitalchat station gave a copy of his previous news, and the above information basically matches, suggesting that the matter has been finalized. He expects OPPO’s own chip to be released in 2024, with subsequent products likely to use a more advanced 3nm design solution.

It is worth mentioning that OPPO has previously launched its own Mariana X and Y chips, while the Mariana Y uses TSMC’s N6RF process and supports Bluetooth 5.3 and LE Audio.

Huawei previously announced a global patent cross-licensing agreement with OPPO Guangdong Mobile Communications Co. In other words, OPPO paid to obtain a license for Huawei’s advanced 5G technology.

In addition, according to the Dongguan release, the OPPO chip R&D center project site was successfully listed on December 27. The project site is located in Dongguan Jiaochengwan Peninsula, with a total investment of 4.5 billion yuan and covers an area of 387 mu, aiming to build a chip R&D center, chip experimental test center, semiconductor equipment research center, 5G terminal R&D center, artificial intelligence R&D center, etc.

According to the investment agreement, the construction period of this project is 36 months, with construction starting by January 2024 at the latest, completion and acceptance by January 2027, and operation by January 2028. After the future production, the annual investment in research and development will be more than 800 million yuan.

The post OPPO phones to use AP chip, will be mass-produced in Q3 next year appeared first on TechGoing.

MediaTek Dimensity 8200 adopts TSMC’s new generation 4nm process technology, with a 1+3+4 architecture design, the performance core is Cortex A78 (*4), the highest frequency is 3.1GHz, plus 4 energy efficiency cores A55, the main frequency is 2.0GHz . Compared with the MediaTek Dimensity 8100, the MediaTek Dimensity 8200 has been upgraded in technology, the CPU frequency has been increased, and the Mali-G610 GPU and APU 580 are integrated at the same time, and the energy efficiency has increased by 8% and 13% respectively. The MediaTek Dimensity 8200 supports FHD+ resolution displays up to 180Hz and WQHD displays up to 120Hz.

In terms of running points, the AnTuTu running score of the mobile phone equipped with Dimensity 8200 can reach 900,000 points, which is much higher than that of MediaTek Dimensity 8100.

MediaTek Dimensity 8200 is also equipped with MediaTek’s Imagiq 785 image processor, which is a 14-bit HDR ISP that provides support for up to 320MP main camera, 4K 60Hz video recording, three cameras, and dual cameras. Exposure for HDR videography. In addition, the chipset uses MediaTek’s APU 580 for artificial intelligence processing, supports 4K AV1 video decoding, adopts 3GPP Release 16 5G mode, supports sub-5GHz 5G networks, supports Bluetooth 5.3 with LE Audio, and Wi-Fi 6E, and also Supports quad-channel LPDDR5 memory and UFS 3.1 storage.

MediaTek Dimensity 8200 supports the Dimensity open architecture, and can cooperate with terminal manufacturers in depth to bring a more personalized and differentiated user experience to terminal devices equipped with Dimensity 8200, and release the powerful potential of the chip.

MediaTek stated that devices using the new Dimensity 8200 SoC will enter the global market starting this month. The iQOO Neo7 SE to be released today will be the world’s first MediaTek Dimensity 8200. In addition, Xiaomi has officially announced the new cup model of the Redmi K60 universe. K60E” is the first batch to be equipped with MediaTek Dimensity 8200.

The post MediaTek Dimensity 8200 processor released: adopts octa-core architecture appeared first on TechGoing.

TSMC will produce 4-nanometer chips at its Arizona plant in 2024, urged by U.S. customers such as Apple, people familiar with the matter said, according to a Bloomberg report last week, adding that the company is expected to announce the new plan when U.S. President Joe Biden visits the plant Tuesday.

TSMC initially planned to produce 20,000 wafers a month in Arizona, but now aims to double its capacity and produce more advanced chips there, the sources said. The initial plan was to build the chips using the same 5nm and 4nm process technology TSMC currently uses to produce the iPhone 14 and iPhone 14 Pro processor chips. Under the new plan, the Arizona facility will eventually produce another 20,000 wafers per month using 3-nanometer technology. Previously TSMC founder Chang Chung-Mou also confirmed that the Arizona plant will also produce 3nm chips, but the timing has not yet been determined.

Sources said that the investment in 3nm expansion may be larger than the $12 billion TSMC invested in the first phase of the plant, but the capacity will be further adjusted according to customer demand.

Earlier, Apple CEO Tim Cook mentioned during a meeting with German employees that “Apple has decided that some of the chips will be sourced from the Arizona factory, which will be operational from 2024. Apple’s latest chips are currently manufactured using the 5nm process, and the shift to more advanced processes will further improve performance and energy efficiency.

Samsung and Intel are also massively expanding their chip production footprint in the United States. Samsung is building a $17 billion chip factory in Texas, while Intel is building a new $20 billion factory in Arizona and another $20 billion factory in Ohio.

The post TSMC’s chip factory in Arizona may start production as early as next year appeared first on TechGoing.

And, a new report says that about a third of the plant’s capacity will be used for Apple. The plant’s previous capacity was said to be about 20,000 full wafers, and the report also says that number could increase over time.
The plan is expected to be fully disclosed on Dec. 6. Tech luminaries including Apple CEO Tim Cook, Nvidia CEO Jen-Hsun Huang and AMD CEO Lisa Su are expected to be on hand. President Joe Biden, and Commerce Secretary Gina Raimondo are also said to be in attendance, according to a Bloomberg report Thursday night.

“We’ve made a decision to buy out capacity from a plant in Arizona, which opens in 2024, so we’ve got about two years on this, maybe a little less,” Cook reportedly said at a conference in Europe in October.” In Europe, I’m sure that as these plans become more apparent, we will also source from Europe if necessary.”

TSMC’s chip manufacturing facility in Arizona begins construction in June 2021. The company initially expected to start production in September 2022, but the schedule was pushed back by about six months – a delay that pushed full chip manufacturing to 2024.

Arizona’s labor pool also reportedly poses a challenge for TSMC. Intel already employs 12,000 people and is seeking to hire another 3,000 for its expanded facility. TSMC will have to compete in an area where unemployment is already low and the availability of clean water may also be an issue when it comes to finding talent for its new plant.

TSMC seems skeptical about the level of success it can achieve in the United States. And, chips purchased from TSMC Arizona will still have to be shipped to China or India for iPhone production if Apple is to keep the current locations for iPhone production intact. A shift to U.S.-based iPhone production by 2024 seems unlikely.

In July, TSMC’s stock fell 35 percent for the year. By Sept. 30, it was down 41 percent. on Nov. 3, it reached its 2022 low, down 49 percent for the year.

The post Apple to purchase one-third of TSMC’s chips produced in Arizona appeared first on TechGoing.

The plant was originally planned to start by producing 5nm chips, but as Apple and other companies look to source parts from the U.S., TSMC has upgraded its plans so that the plant can offer more advanced chips.

TSMC is expected to announce the new plans next Tuesday, sources said. The sources said Apple will use 4nm and 3nm processes for its upcoming M-series and A-series chips for the Mac, iPad, iPhone and other products.

TSMC responded last month to the news of a “$12 billion investment in a new 3nm plant in the United States” by saying that it has not yet determined the plans for the second phase of the Arizona fab. “Given the strong customer demand for the company’s advanced processes, we will evaluate future plans in terms of operational efficiency and cost economics.”

The post At Apple’s request, sources say TSMC’s Arizona plant to start producing 4nm chips in 2024 appeared first on TechGoing.